Studies of adjunctive therapy with corticosteroids have suggested benefits for patients with community-acquired pneumonia (CAP). However, pointing to a “lack of unequivocal conclusions…on hard clinical endpoints,” researchers from Geneva University Hospitals and Geneva Faculty of Medicine in Switzerland decided to do an updated meta-analysis of risks and advantages. They reviewed 14 trials involving 2,077 patients who were treated with either adjunctive corticotherapy or antimicrobial therapy.
Eight studies, with 1,624 patients, reported on length of stay. They found a mean of 9 days for patients treated with adjunctive corticotherapy, vs 10.6 days for those treated with antimicrobial therapy alone.
Four studies, with 1,163 patients, reported on time to clinical stability (TCS). The mean TCS was 3.3 days for patients on adjunctive corticotherapy compared with 4.3 days for antimicrobial treatment alone.
Four studies with 304 patients reported on severe complications that required vasopressors or mechanical ventilation. In those, 3 times as many patients on antimicrobial therapy had severe complications (37.4% vs 12.1%). In 4 other studies, 3.6% of 138 patients on adjunctive corticosteroids needed vasopressors, compared with 14.5% of 128 patients on antimicrobial therapy. And in 7 studies reporting on mechanical ventilation, 5.3% of 619 adjunctive corticosteroid patients required ventilation compared with 12.1% of 580 on antimicrobials. Although mortality was significantly reduced in a subgroup of patients with severe CAP, the researchers found no benefit in 30-day mortality.
Adverse events seem to be infrequent and short lasting, the researchers say, although the risk of hyperglycemia was higher among patients on corticosteroids. Adjunctive therapy increased the absolute proportion of patients who required insulin therapy during hospitalization by 7.2%, but the number of patients with new insulin dependence at day 30 was “very low.” Glucocorticoids may have some deleterious effects after treatment withdrawal, the researchers say, but their meta-analysis did not have enough data to confirm those concerns.
Source:
Marti C, Grosgurin O, Harbarth S, et al. PLoS ONE. 2015;10(12):e0144032.
doi:10.1371/journal.pone.0144032.